Fault diagnosis method for shared DC power supply system in substation

Abstract

The present invention discloses a fault diagnosis method for a substation shared DC power supply system, which relates to the field of intelligent processing of electric power data. The method includes: obtaining topological structure information of a shared battery pack and establishing an initial weight; reading real-time work tasks, applying analysis to the topological structure information, and adjusting the initial weight; performing data monitoring and establishing a monitoring data set; calculating health status, calculating abnormal correlation, and generating component health indicators; inputting a health evaluation network and performing comprehensive health indicator calculations; configuring monitoring verification nodes according to the comprehensive health indicator calculation results and target work tasks, performing additional monitoring, and generating fault diagnosis results according to the additional monitoring results. The method solves the technical problem of low diagnostic accuracy and reliability of fault diagnosis of shared battery packs in existing substation shared DC power supply systems, and achieves the technical effect of improving the accuracy and reliability of fault diagnosis.

Description

Technical Field

[0001] The present application relates to the field of intelligent processing of electric power data, and in particular to a fault diagnosis method for a shared direct current power supply system of a substation. Background Art

[0002] As power systems become increasingly complex and intelligent, the importance of substation shared DC power systems as a key link in ensuring the stable operation of power facilities is becoming increasingly prominent. Substation shared DC power systems integrate multiple power function modules such as DC power supply, AC uninterruptible power supply, power inverter and communication power supply to share one or more battery packs, thereby greatly reducing the duplication of battery configuration, reducing costs and floor space. However, in terms of health monitoring of shared battery packs, traditional monitoring methods often use models based on a single data source for health assessment, lacking in-depth analysis of the correlation between abnormal patterns, resulting in an inability to accurately determine the cause and scope of the abnormality.

[0003] In the current related technologies, fault diagnosis of shared battery packs in a substation shared DC power supply system has technical problems of low diagnostic accuracy and reliability. Summary of the invention

[0004] The present application provides a fault diagnosis method for a shared DC power supply system in a substation. The method connects to the shared DC power supply system in the substation, obtains the topological structure information of the shared battery pack, analyzes the topological structure information according to real-time work tasks, dynamically adjusts the weights of components in the shared battery pack, establishes a monitoring data set including multivariate data such as voltage, current, temperature, and resistance, uses the monitoring data set to calculate the health status and abnormal correlation of the components, generates component health indicators, inputs the components into a health evaluation network to calculate the comprehensive health indicators, configures monitoring verification nodes based on the calculation results, performs additional monitoring and other technical means, thereby achieving comprehensive, accurate and reliable fault diagnosis of the shared battery pack, and achieving the technical effect of improving the accuracy and reliability of fault diagnosis.

[0005] The present application provides a fault diagnosis method for a substation shared DC power supply system, comprising:

[0006] Connecting to the substation shared DC power supply system, obtaining the topological structure information of the shared battery group, and establishing the initial weights of the components in the shared battery group according to the topological structure information; reading the real-time work tasks of the substation shared DC power supply system, performing application analysis on the topological structure information according to the real-time work tasks, adjusting the initial weights of the components in the shared battery group, and generating application weights; synchronously performing data monitoring of the components in the shared battery group, and establishing a monitoring data set consisting of voltage data, current data, temperature data, and resistance data; calculating the health status of the components in the shared battery group through the monitoring data set, and calculating the abnormal correlation through the application weights, and generating component health indicators according to the abnormal correlation and health status; inputting the component health indicators into the health evaluation network, calculating the comprehensive health indicators of the shared battery group, and generating comprehensive health indicator calculation results, wherein the comprehensive health indicator calculation process includes analyzing the abnormal correlation of components in different shared battery groups; configuring a monitoring verification node according to the comprehensive health indicator calculation results and the target work tasks, performing additional monitoring through the monitoring verification node, and generating a fault diagnosis result of the shared battery group according to the additional monitoring results.

[0007] In a possible implementation, the component health indicator is generated according to the abnormal correlation and the health status, and the following processing is performed:

[0008] The calculation formula for component health indicators is established as follows: ;in, Indicates components in a shared battery pack Component health indicators, Indicates components in a shared battery pack The health status weight, Indicates components in a shared battery pack health status, is the influence coefficient determined by the abnormal correlation degree, Different from the components in the shared battery pack Any shared battery pack components, It is a shared battery pack component Shared battery pack components The abnormal correlation degree of the component is calculated according to the calculation formula to obtain the component health index.

[0009] In a possible implementation, the component health index is input into a health evaluation network, a comprehensive health index of a shared battery pack is calculated, a comprehensive health index calculation result is generated, and the following processing is performed:

[0010] The health evaluation network is as follows: ;in, Characterize the calculation results of comprehensive health indicators, is the total number of components in the shared battery pack, Characterizing components within a shared battery pack The component health index influences the weight, is the adjustment factor used to characterize the components in different shared battery packs and shared battery pack components The importance of the combined abnormal mode; and calculating the comprehensive health index of the shared battery pack based on the health evaluation network.

[0011] In a possible implementation, the following processing is performed:

[0012] A backtracking window is configured, and backtracking monitoring and verification of the shared battery group is performed through the backtracking window, and a backtracking feature set is configured; the abnormal correlation is corrected through the backtracking feature set, and the influence coefficient is established.

[0013] In a possible implementation, the abnormal correlation is corrected by the backtracking feature set, the influence coefficient is established, and the following processing is performed:

[0014] Acquire the time mark of the retrospective feature set and establish an initial time series attenuation factor; perform feature analysis on the retrospective feature set to establish a feature basic influence strength; compensate the feature basic influence strength through the initial time series attenuation factor to construct a time series optimization coefficient; correct the abnormal correlation based on the time series optimization coefficient and configure the dynamically updated influence coefficient.

[0015] In a possible implementation, the component health indicator is input into a health assessment network, and the following processing is performed:

[0016] Call the preprocessing subnetwork in the health assessment network to perform preprocessing of component health indicators and generate a preprocessing data set; synchronize the preprocessing data set to the independent analysis subnetwork, perform independent abnormal analysis of components in the shared battery pack based on the independent analysis subnetwork, and generate a first health indicator calculation result according to the independent abnormal analysis result; obtain abnormal features in the independent abnormal analysis, and configure abnormal associations based on the abnormal features; synchronize the abnormal associations to the linkage analysis subnetwork, and after receiving the preprocessing data set through the linkage analysis subnetwork, perform linkage abnormal analysis, and generate a second health indicator calculation result according to the linkage abnormal analysis result; generate a comprehensive health indicator calculation result according to the first health indicator calculation result and the second health indicator calculation result.

[0017] In a possible implementation, the following processing is performed:

[0018] The independent abnormal analysis results are compared with corresponding abnormal thresholds to generate abnormal transition results; and component abnormalities of components in the shared battery pack are reported based on the abnormal transition results.

[0019] In a possible implementation, the following processing is performed:

[0020] Establish a life cycle of a shared battery pack and configure a tolerance penalty coefficient mapped to the life cycle; adaptively update the abnormal correlation through the tolerance penalty coefficient, and reconstruct the component health indicator with the adaptive update result.

[0021] The fault diagnosis method for the substation shared DC power supply system proposed in this application first connects the substation shared DC power supply system, obtains the topological structure information of the shared battery group, establishes the initial weights of the components in the shared battery group according to the topological structure information, then reads the real-time work tasks of the substation shared DC power supply system, performs application analysis on the topological structure information according to the real-time work tasks, adjusts the initial weights of the components in the shared battery group, generates application weights, and then synchronously performs data monitoring of the components in the shared battery group, establishes a monitoring data set consisting of voltage data, current data, temperature data, and resistance data, and then calculates the health status of the components in the shared battery through the monitoring data set, and calculates the abnormal correlation by applying the weight, generates component health indicators according to the abnormal correlation and health status, and then inputs the component health indicators into the health evaluation network, calculates the comprehensive health indicators of the shared battery group, and generates comprehensive health indicator calculation results. The comprehensive health indicator calculation process includes analyzing the abnormal correlation of components in different shared battery groups. Finally, the monitoring verification node is configured according to the comprehensive health indicator calculation results and the target work tasks, and additional monitoring is performed through the monitoring verification node. The fault diagnosis results of the shared battery group are generated according to the additional monitoring results, thereby achieving the technical effect of improving the accuracy and reliability of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the accompanying drawings of the embodiment of the present invention will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the method according to the embodiment of the present application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0023] Figure 1 A flowchart of a fault diagnosis method for a substation shared DC power supply system provided in an embodiment of the present application.

[0024] Figure 2 A schematic diagram of a flow chart for generating comprehensive health index calculation results in a fault diagnosis method for a substation shared DC power supply system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0025] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0026] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0027] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product, or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0028] The present application embodiment provides a fault diagnosis method for a substation shared DC power supply system, such as Figure 1 As shown, the method includes:

[0029] Step S100, connect to the substation shared DC power supply system, obtain the topological structure information of the shared battery pack, and establish the initial weights of the components in the shared battery pack according to the topological structure information. Specifically, determine the communication interface (such as Ethernet, serial port, etc.) and communication protocol (such as Modbus, IEC 61850, etc.) between the substation shared DC power supply system, and establish a physical connection and communication connection with the substation shared DC power supply system through the configured interface and protocol. Through the established connection, send a request to obtain the topological structure information of the shared battery pack to the substation shared DC power supply system, receive the returned topological structure information data, including the number of batteries, connection mode (series, parallel, etc.), protection equipment (such as fuses, circuit breakers) configuration, etc., and parse the received data into a topological structure diagram or model. Analyze the resolved topology diagram or model to determine the position and role of each component in the system. Determine the basis for weight allocation based on factors such as component importance, failure rate, and impact on the system. Based on the determined weight factors, assign an initial weight value to each component (such as a single battery, fuse, etc.). The weight value can be a numerical value or a level, which is used to indicate the importance of the component in the system and the impact of the failure.

[0030] Step S200, read the real-time work task of the substation shared DC power supply system, perform application analysis on the topology information according to the real-time work task, adjust the initial weight of the components in the shared battery pack, and generate the application weight. Specifically, through the established communication connection, send a request to the substation shared DC power supply system to obtain the current real-time work task information, and receive the returned real-time work task data, including load size, working time, working mode (such as charging, discharging, standby, etc.). Based on the specific content of the real-time work task, analyze the impact of the real-time work task on each component in the shared battery pack. For example, a high-load task will cause the battery pack temperature to rise, thereby affecting the battery life. According to the task requirements, evaluate whether each component is in a normal working state and whether there are potential risks or hidden dangers. Combine the real-time work task with the topology information, analyze the impact of the real-time work task on the overall structure of the shared battery pack, and the role and importance of each component in the task execution. According to the results of the application analysis, determine the components whose weights need to be adjusted and the basis for the adjustment (such as the degree of impact of the task on the component, the failure rate of the component, etc.). According to the determined adjustment factors, calculate the adjustment value of each component weight, and the adjustment value can be an increased or decreased value or percentage. Apply the calculated adjustment value to the initial weight to obtain a new weight value, i.e., the applied weight.

[0031] Step S300, synchronously perform data monitoring of components in the shared battery pack, and establish a monitoring data set consisting of voltage data, current data, temperature data, and resistance data. Specifically, deploy voltage, current, temperature, and resistance sensors on key components in the shared battery pack (such as single cells, busbars, protection devices, etc.), and configure the data acquisition system (such as data acquisition card, PLC, etc.) according to the type of sensor, communication method, and acquisition frequency. Through the data acquisition system, each sensor is synchronously started for data acquisition, so that all sensors start collecting data at the same time. The raw data collected from the sensor is sorted, including data cleaning (removing abnormal values, duplicate values, etc.), data conversion (converting electrical signals into actual physical values), etc. The sorted data is classified according to types such as voltage, current, temperature, and resistance, and stored in corresponding data sets respectively, and each classified data set is integrated to form a comprehensive monitoring data set containing voltage data, current data, temperature data, and resistance data.

[0032] Step S400, calculate the health status of the components in the shared battery pack through the monitoring data set, calculate the abnormal correlation through the application weight, and generate the component health index according to the abnormal correlation and health status. Specifically, extract the voltage, current, temperature, resistance and other parameter data related to a specific component (such as a single battery) from the monitoring data set, and analyze the current status of the component according to the extracted data. For example, by comparing the voltage and current data with the standard value or historical data, judge whether the battery is in a normal discharge or charging state; judge whether the battery is overheated through temperature data; judge whether the internal resistance of the battery increases through resistance data, etc. Based on the results of parameter analysis, an algorithm or model (such as a GMDH-based model, SOH calculation formula, etc.) is used to calculate the health status of the component. The health status refers to the current working state and health level of the component, which can be expressed by a value or level, such as a score of 0-100 or A, B, C levels.

[0033] Based on factors such as the physical connection between components, working mode, and historical fault data, the abnormal correlation between different components is analyzed. According to the application weight, a weight value is assigned to each associated factor. The weight value reflects the degree of influence of each factor on the health status of the component. Based on the associated factors and the assigned weight values, the abnormal correlation is calculated. The abnormal correlation indicates the degree of influence of the abnormal state of one component on another component, which can be a numerical value.

[0034] The health status and abnormal correlation of the component are comprehensively evaluated to assess the overall health status of the component. Based on the results of the comprehensive evaluation, the health indicator of the component is generated. The component health indicator can be a value, a level, or a descriptive text to intuitively represent the health status of the component.

[0035] In a possible implementation, the component health index is generated according to the abnormal correlation and the health status, and step S400 further includes step S410, establishing a calculation formula for the component health index, as follows:

[0036] ;

[0037] in, Indicates components in a shared battery pack Component health indicators, Indicates components in a shared battery pack The health status weight, Indicates components in a shared battery pack health status, is the influence coefficient determined by the abnormal correlation degree, Different from the components in the shared battery pack Any shared battery pack components, It is a shared battery pack component Shared battery pack components abnormal correlation; step S420, calculate and obtain the component health index according to the calculation formula. Specifically, the health status weight represents the importance of the component in the shared battery pack, the health status is the current health status or performance evaluation of the component, the abnormal correlation is the degree of influence of the abnormal state of one component on another component, and the influence coefficient is a dynamic coefficient determined according to the abnormal correlation and used to calculate the health index. Collect the health status data, health status weight, abnormal correlation with other components and other information of each component, set an influence coefficient for each component on other components according to the collected abnormal correlation data, substitute the collected data into the calculation formula established in step S410, and calculate the health index of each component. This implementation method takes into account the importance of the component in the shared battery pack, its own health status and the abnormal correlation with other components, so as to achieve the technical effect of comprehensively evaluating the health status of the component.

[0038] In a possible implementation, step S410 further includes step S411, configuring a backtracking window, performing backtracking monitoring and verification of the shared battery pack through the backtracking window, and configuring a backtracking feature set. Specifically, based on the operating characteristics, fault frequency, and data analysis requirements of the shared battery pack, the time range of the backtracking window, i.e., the historical time period to be backtracked and monitored, is determined. Within the selected backtracking window, the historical operating data of the shared battery pack is monitored and verified, and key features are extracted from the historical data in the backtracking window, including operating parameters, fault records, environmental parameters, etc. of the components, and the extracted features are organized into a feature set according to certain rules and logic, i.e., the backtracking feature set. Step S412, the abnormal correlation is corrected through the backtracking feature set, and the influence coefficient is established. Specifically, the features in the backtracking feature set are analyzed by statistical analysis, machine learning, and other methods to determine how these features affect the abnormal correlation of the components. According to the analysis results of the backtracking feature set, the abnormal correlation is corrected, including adjusting the value of the abnormal correlation, etc. Based on the corrected abnormal correlation, an influence coefficient is set for the influence of each component on other components, which is used to calculate the component health index. This implementation method more accurately evaluates the abnormal correlation between components through retrospective monitoring verification and feature analysis, and establishes the influence coefficient accordingly. The introduction of the retrospective window enables the use of historical data to enhance the accuracy and reliability of the analysis. By correcting the abnormal correlation by retrospective feature sets, the mutual influence relationship between components is more finely portrayed, thereby more accurately evaluating the health status of components, achieving the technical effect of improving the accuracy and practicality of component health indicator calculation.

[0039] In a possible implementation, the abnormal correlation is corrected by the retrospective feature set to establish the influence coefficient, and step S412 further includes step S4121, obtaining the time mark of the retrospective feature set and establishing an initial time series attenuation factor. Specifically, the time mark of each feature is extracted from the retrospective feature set, including the time of event occurrence. According to the time mark, an initial time series attenuation factor is set for each feature, and the initial time series attenuation factor is a time-related function or value, which is used to represent the influence of feature data decaying over time. Step S4122, feature analysis is performed on the retrospective feature set to establish the basic influence strength of the feature. Specifically, features closely related to the abnormal correlation analysis are selected from the retrospective feature set, and statistical analysis, machine learning or other related analysis are performed on the selected features to evaluate the impact on the abnormal correlation. According to the result of the feature analysis, a basic influence strength is determined for each feature, and the basic influence strength of the feature represents the direct influence of the feature on the abnormal correlation, which can be a value. Step S4123, the basic influence strength of the feature is compensated by the initial time series attenuation factor to construct a time series optimization coefficient. Specifically, the initial time series attenuation factor is applied to the feature basic influence strength, time attenuation compensation is performed, and a time series optimization coefficient is calculated by combining the influence strength after attenuation compensation and the original basic influence strength. The time series optimization coefficient is a coefficient that combines the feature basic influence strength and the time attenuation effect, and is used to optimize the calculation of the abnormal correlation. Step S4124, the abnormal correlation is corrected based on the time series optimization coefficient, and the dynamically updated influence coefficient is configured. Specifically, the abnormal correlation is corrected using the time series optimization coefficient to more accurately reflect the influence of the feature on the abnormal correlation. According to the corrected abnormal correlation, a dynamically updated influence coefficient is configured for each component, and this influence coefficient is dynamically adjusted with the change of time and feature data. This implementation method more accurately reflects the influence of feature data decaying with time by introducing a time series attenuation factor and a time series optimization coefficient, and corrects the abnormal correlation accordingly. Since the operating state of a device or system changes over time, the old data has little influence on the current state. Therefore, through time attenuation compensation, the influence of historical data on the current abnormal correlation is more reasonably evaluated, and the technical effect of making the abnormal correlation evaluation result more accurate and reliable is achieved.

[0040] In a possible implementation, step S400 further includes step S430, establishing the life cycle of the shared battery pack and configuring a tolerance penalty coefficient mapped to the life cycle. Specifically, collect relevant data of the shared battery pack, such as the number of charges, discharge depth, temperature history, etc., and analyze and predict the life cycle of the shared battery pack based on the collected data through statistical analysis, machine learning or physical models. The life cycle is the total time or total number of cycles from a brand new state to a significant performance degradation (such as a capacity drop below a certain threshold) of the shared battery pack. According to the life cycle of the shared battery pack, configure a tolerance penalty coefficient mapped to the life cycle. The tolerance penalty coefficient is a function that changes with time or the number of cycles, which is used to represent the tolerance of the shared battery pack performance decay over time. For example, when the shared battery pack approaches the end of its life cycle, the tolerance penalty coefficient will increase to reflect the impact of the performance degradation of the shared battery pack on the system. Step S440, adaptively update the abnormal correlation through the tolerance penalty coefficient, and reconstruct the component health indicator with the adaptive update result. Specifically, the tolerance penalty coefficient is applied to the abnormal correlation and adaptively updated. That is, according to the current life cycle status of the shared battery pack (that is, how close it is to the end of its life cycle), the value of the abnormal correlation is adjusted. Based on the updated abnormal correlation, the health indicators of the components are re-evaluated and constructed. This implementation method more accurately evaluates the current status and potential problems of the shared battery pack by configuring the tolerance penalty coefficient mapped with the life cycle and updating the abnormal correlation accordingly, thereby achieving the technical effect of providing more accurate component health indicators and improving the accuracy and reliability of fault diagnosis.

[0041] Step S500, input the component health index into the health evaluation network, calculate the comprehensive health index of the shared battery group, and generate the comprehensive health index calculation result. The comprehensive health index calculation process includes analyzing the abnormal association of components in different shared battery groups. Specifically, collect and organize the health indicators of each component in the generated shared battery group, and import the organized component health indicators as input data into the pre-built health evaluation network (a computing model or system for evaluating the health status of the shared battery group). The health evaluation network pre-processes the input component health indicators, such as data standardization and normalization, analyzes the abnormal association between components in different shared battery groups based on the connection relationship, working mode and historical data between components, and calculates the abnormal propagation probability or impact degree between components through algorithms (such as neural networks, graph algorithms, etc.). Based on the component health indicators and the abnormal association analysis results, the health evaluation network uses algorithms (such as weighted average, machine learning models, etc.) to perform comprehensive calculations to obtain the comprehensive health indicators of the shared battery group. The comprehensive health indicator can be a numerical value, level or descriptive text that describes the overall health status of the shared battery group, which integrates the health status of each component in the shared battery group and the abnormal association between components.

[0042] In a possible implementation, the component health index is input into a health evaluation network, a comprehensive health index of the shared battery pack is calculated, and a comprehensive health index calculation result is generated. Step S500 further includes step S510, and the health evaluation network is as follows:

[0043] ;

[0044] in, Characterize the calculation results of comprehensive health indicators, is the total number of components in the shared battery pack, Characterizing components within a shared battery pack The component health index influences the weight, is the adjustment factor used to characterize the components in different shared battery packs and shared battery pack components Importance of the combined abnormal mode; Step S520, calculate the comprehensive health index of the shared battery pack according to the health evaluation network. Specifically, the health evaluation network is used to integrate the health indicators of each component in the shared battery pack and calculate the comprehensive health index. The calculation result of the comprehensive health index is a value or level that can reflect the overall health status of the shared battery pack. The component health indicator data is input into the health evaluation network, and the health evaluation network calculates the comprehensive health index of the shared battery pack based on the input component health indicators and other parameters. This implementation method more accurately calculates the comprehensive health index of the shared battery pack by integrating the health indicators of multiple components and taking into account the influence weights of different components and the adjustment coefficients under specific abnormal modes, thereby achieving the technical effect of more accurately evaluating the overall health status of the shared battery pack.

[0045] like Figure 2As shown, in a possible implementation, the component health indicator is input into the health evaluation network, and step S500 further includes step S530, calling the preprocessing sub-network in the health evaluation network to perform preprocessing of the component health indicator and generate a preprocessing data set. Specifically, the collected raw health indicator data is cleaned, including processing abnormal values, duplicate data, etc., and the cleaned data is converted into a form suitable for processing by the health evaluation network, which may include normalization, standardization, encoding and other steps, selecting features useful for health evaluation, or extracting new features from the raw data, and organizing the preprocessed data into a preprocessing data set, ready for subsequent analysis steps. Step S540, synchronize the preprocessing data set to the independent analysis sub-network, perform independent abnormal analysis of the components in the shared battery pack based on the independent analysis sub-network, and generate a first health indicator calculation result according to the independent abnormal analysis result. Specifically, the preprocessed data set is transmitted to the independent analysis sub-network, which is a sub-module in the health evaluation network, and is used to independently analyze the health indicators of a single component and identify abnormalities or potential problems. Based on the results of the independent abnormal analysis, a preliminary health indicator calculation result is generated for each component. Step S550, obtain the abnormal features in the independent abnormal analysis, and configure the abnormal association based on the abnormal features. Specifically, extract the key abnormal features from the results of the independent abnormal analysis, analyze the relationship between the abnormal features, and determine the degree of association between the abnormal features. Step S560, synchronize the abnormal association to the linkage analysis subnetwork, and after receiving the preprocessed data set through the linkage analysis subnetwork, perform linkage abnormal analysis, and generate a second health indicator calculation result according to the linkage abnormal analysis result. Specifically, transmit the configured abnormal association information to the linkage analysis subnetwork, which is a submodule in the health evaluation network, and is used to consider the abnormal association between different components and perform comprehensive analysis. Based on the results of the linkage abnormal analysis, generate a more comprehensive and accurate health indicator calculation result. Step S570, generate a comprehensive health indicator calculation result based on the first health indicator calculation result and the second health indicator calculation result. Specifically, the first health indicator calculation result and the second health indicator calculation result are integrated, and the two results can be weighted and adjusted as needed to reflect the different importance of the two in the comprehensive health indicator, and the final comprehensive health indicator calculation result is generated based on the results after integration and weighted adjustment. This implementation method comprehensively and accurately evaluates the health status of the shared battery pack by introducing a preprocessing subnetwork, an independent analysis subnetwork, and a linkage analysis subnetwork. The preprocessing subnetwork ensures the quality and consistency of the input data; the independent analysis subnetwork identifies anomalies or potential problems of a single component; and the linkage analysis subnetwork considers the abnormal correlation between different components and conducts a more comprehensive and accurate analysis.By integrating the two analysis results, a more comprehensive and accurate calculation result of the comprehensive health index is obtained, achieving the technical effect of providing a more reliable basis for the maintenance and management of shared battery packs.

[0046] In a possible implementation, step S540 further includes step S541, comparing the independent abnormal analysis result with the corresponding abnormal threshold value to generate an abnormal transition result. Specifically, for the health indicators of each component in the shared battery pack, one or more abnormal thresholds are pre-set based on the normal working range of the component, historical data, expert experience and other factors. In the independent analysis subnetwork, the health indicators of each component are independently analyzed to identify possible abnormalities or potential problems, and the results of the independent abnormal analysis are compared with the pre-set abnormal thresholds. If the health indicator of a component exceeds the set threshold range, it is considered that the component has undergone an abnormal transition. According to the result of the threshold comparison, an abnormal transition result is generated, including the component identification, abnormal type, abnormal degree and other information of the abnormal transition. Step S542, the component abnormality of the component in the shared battery pack is reported through the abnormal transition result. Specifically, the abnormal transition result is parsed to determine the component that has undergone abnormal transition and its related information, and a component abnormality report is generated according to the parsed result, including the component identification, abnormal type, abnormal degree, abnormal time and other information of the abnormal component. This implementation method accurately identifies which components' health indicators exceed the normal range by comparing independent abnormal analysis results with abnormal thresholds, thereby generating abnormal transition results. Component abnormalities are reported through abnormal transition results, and detailed information about abnormal components is provided to relevant personnel in a timely manner so that timely measures can be taken to repair or replace them. This achieves the technical effect of ensuring the normal operation and safety of shared battery packs and reducing potential risks caused by component abnormalities.

[0047] Step S600, configure a monitoring verification node according to the calculation result of the comprehensive health index and the target work task, perform additional monitoring through the monitoring verification node, and generate a fault diagnosis result of the shared battery pack according to the additional monitoring result. Specifically, according to the calculation result of the comprehensive health index and the target work task, determine the specific area or component that needs additional monitoring, as well as the requirements such as the frequency and accuracy of monitoring. According to the monitoring requirements, select a suitable monitoring verification node for configuration. The monitoring verification node can be a pre-set fixed node or a mobile node temporarily deployed according to the requirements of the target work task. The configuration of the monitoring verification node includes determining the location, connection method and monitoring parameters of the node. According to the configuration requirements, start the monitoring verification node to collect real-time operating data of the shared battery pack to verify or confirm the existence and location of the fault. The monitoring verification node transmits the collected data to the central processing system through a preset communication method (such as a wired network, a wireless network, etc.). The central processing system analyzes the received additional monitoring data, identifies the abnormality or fault mode therein, and uses a fault diagnosis algorithm (such as rule-based reasoning, machine learning, etc.) based on the results of the data analysis to perform fault diagnosis on the shared battery pack. The diagnosis results include the fault type, fault location and possible fault cause. The fault diagnosis results are output to users or related systems in the form of reports, charts, etc., so that corresponding maintenance or repair measures can be taken. The embodiment of the present application adopts a shared DC power supply system connected to a substation, obtains the topological structure information of a shared battery pack, analyzes the topological structure information according to real-time work tasks, dynamically adjusts the weights of components in the shared battery pack, establishes a monitoring data set including multivariate data such as voltage, current, temperature, and resistance, uses the monitoring data set to calculate the health status and abnormal correlation of the components, generates component health indicators, inputs the health evaluation network to calculate the comprehensive health indicator, configures the monitoring verification node based on the calculation results, performs additional monitoring and other technical means, realizes comprehensive, accurate, and reliable fault diagnosis of shared battery packs, and achieves the technical effect of improving the accuracy and reliability of fault diagnosis.

[0048] The above specific implementation manner does not constitute a limitation to the protection scope of the present application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application. In some cases, the actions or steps recorded in the present application can be performed in an order different from that in the embodiment and can still achieve the desired results. In addition, the process depicted in the accompanying drawings does not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A fault diagnosis method for a substation shared DC power supply system, characterized in that: The method comprises: Connecting to the substation shared DC power supply system, obtaining topological structure information of the shared battery group, and establishing initial weights of components in the shared battery group according to the topological structure information; Reading the real-time working tasks of the substation shared DC power supply system, performing application analysis on the topology information according to the real-time working tasks, adjusting the initial weights of the components in the shared battery group, and generating application weights; Synchronously perform data monitoring of components in a shared battery pack and establish a monitoring data set consisting of voltage data, current data, temperature data, and resistance data; Calculate the health status of the components in the shared battery pack through the monitoring data set, calculate the abnormal correlation through the application weight, and generate the component health index according to the abnormal correlation and the health status; Inputting the component health index into a health evaluation network, calculating a comprehensive health index of the shared battery group, and generating a comprehensive health index calculation result, wherein the comprehensive health index calculation process includes analyzing abnormal correlations between components in different shared battery groups; According to the calculation result of the comprehensive health index and the target work task, a monitoring verification node is configured, additional monitoring is performed through the monitoring verification node, and a fault diagnosis result of the shared battery pack is generated according to the additional monitoring result; The generating of component health indicators according to the abnormal correlation and health status further includes: The calculation formula for component health indicators is established as follows: ； in, Indicates components in a shared battery pack Component health indicators, Indicates components in a shared battery pack The health status weight, Indicates components in a shared battery pack health status, is the influence coefficient determined by the abnormal correlation degree, Different from the components in the shared battery pack Any shared battery pack components, It is a shared battery pack component Shared battery pack components The abnormal correlation degree; Calculate and obtain component health indicators according to the calculation formula; The step of inputting the component health index into a health evaluation network, calculating a comprehensive health index of a shared battery pack, and generating a comprehensive health index calculation result further includes: The health evaluation network is as follows: ； Among them, the calculation results of comprehensive health indicators are represented. is the total number of components in the shared battery pack, Characterizing components within a shared battery pack The component health indicator influences the weight, is the adjustment factor used to characterize the components in different shared battery packs and shared battery pack components The importance of the combination of abnormal patterns; The comprehensive health index of the shared battery pack is calculated according to the health evaluation network.

2. The fault diagnosis method for a substation shared DC power supply system according to claim 1, characterized in that: The method further comprises: Configure a backtracking window, perform backtracking monitoring verification of the shared battery group through the backtracking window, and configure a backtracking feature set; The abnormal correlation is corrected through the backtracking feature set to establish the influence coefficient.

3. The fault diagnosis method for a substation shared DC power supply system according to claim 2, characterized in that: The correction of the abnormal correlation degree by the backtracking feature set to establish the influence coefficient also includes: Obtain the time stamp of the backtracking feature set and establish the initial time series decay factor; Performing feature analysis on the backtracking feature set to establish feature basic influence strength; The initial time series attenuation factor is used to compensate the characteristic basic influence intensity, and a time series optimization coefficient is constructed; The abnormal correlation is modified based on the timing optimization coefficient, and the dynamically updated influence coefficient is configured.

4. The fault diagnosis method for a substation shared DC power supply system according to claim 1, characterized in that: The step of inputting the component health indicator into a health assessment network further includes: Calling the preprocessing sub-network in the health evaluation network to perform preprocessing of component health indicators and generate a preprocessing data set; Synchronizing the preprocessed data set to an independent analysis sub-network, performing independent abnormality analysis of components in the shared battery pack based on the independent analysis sub-network, and generating a first health indicator calculation result according to the independent abnormality analysis result; Acquire abnormal features in independent abnormal analysis, and configure abnormal association based on the abnormal features; Synchronizing the abnormal association to the linkage analysis sub-network, and after receiving the pre-processed data set through the linkage analysis sub-network, performing linkage abnormal analysis, and generating a second health indicator calculation result according to the linkage abnormal analysis result; A comprehensive health index calculation result is generated according to the first health index calculation result and the second health index calculation result.

5. The fault diagnosis method for a substation shared DC power supply system according to claim 4, characterized in that: The method further comprises: Performing a corresponding abnormal threshold comparison on the independent abnormal analysis results to generate an abnormal transition result; The abnormal transition result is used to report the abnormality of the components in the shared battery pack.

6. The fault diagnosis method for a substation shared DC power supply system according to claim 1, characterized in that: The method further comprises: Establish the life cycle of the shared battery pack and configure the tolerance penalty coefficient mapped with the life cycle; The abnormal correlation is adaptively updated through the tolerance penalty coefficient, and the component health index is reconstructed with the adaptive update result.

Images